COMMENT AND DEBATE COMMENT AND DEBATE What can the life course approach contribute to an understanding of longevity risk? David Blane ESRC International Centre for Life Course Studies in Society Health and University College London, UK [email protected] Bola Akinwale ESRC International Centre for Life Course Studies in Society and Health and Imperial College London, UK Rebecca Landy ESRC International Centre for Life Course Studies in Society and Health and Queen Mary College London, UK Katey Matthews ESRC International Centre for Life Course Studies in Society and Health and University of Manchester, UK Morten Wahrendorf ESRC International Centre for Life Course Studies in Society and Health and University of Duesseldorf, Germany Hans-Werner Wahl Heidelberg University, Germany Mark D. Hayward University of Texas at Austin, USA Aart C. Liefbroer Netherlands Interdisciplinary Demographic Institute, University Medical Center Groningen and Vrije Universiteit Amsterdam, Netherlands Gita D. Mishra University of Queensland, Australia Isabel Ferreira University of Queensland, Australia Ilona Koupil Stockholm University Karolinska Institute, Sweden http://dx.doi.org/10.14301/llcs.v7i2.343 Abstract Longevity risk means living longer than predicted. Attempts to understand longevity risk to date have concentrated on single diseases, usually coronary heart disease, and sought explanations in terms of risk factor change and medical innovation. In an opening paper, David Blane and colleagues point to evidence that suggests changes in positive health also should be considered; and that a life course approach can do so in a way that is socially and biologically plausible. Applying this approach to UK citizens currently aged 85 years suggests that life course research should give priority to trajectories across the whole life course and to the social and material contexts through which each cohort has passed. Testing these ideas will require inter-disciplinary and international comparative research. The opening paper is followed by commentaries from Hans-Werner Wahl, Mark Hayward, Aart Liefbroer and Gita Mishra. Finally Blane and colleagues respond to the points raised by the commentators. Keywords Longevity risk, positive health, social and biological plausibility, life course trajectories, social history context 165 COMMENT AND DEBATE Introduction or the introduction of new medical treatments. The increase in life expectancy at middle age was Positive health, in the sense of growth and one of the defining characteristics of the late development, functional capacity, vitality and twentieth century, with all-cause mortality around resilience, was largely ignored. The present paper the state pension age in England and Wales falling by attempts to rectify this neglect by arguing that some two thirds during 1971-2001 (Akinwale et al., positive health should be considered alongside risk 2011), and similar improvements occurring in many factor change and medical innovation as driving the other countries (World Health Organisation, 2008). increase in life expectancy at middle age; and that The actuarial profession led research into the causes the life course approach can contribute to the of this change. Their thinking includes the idea of the understanding of longevity risk by bringing both golden cohort, born in the 1930s, who were the first biological and social plausibility (Blane, Kelly-Irving, in the UK to experience the fall in middle aged d’Errico, Bartley & Montgomery, 2013) to the idea of mortality (Willets, 2004), and the concept of positive health. longevity risk to describe the phenomenon of living Two pieces of evidence support the idea of positive longer than predicted (Willets et al., 2004). health. Table 1 shows for England and Wales, the The rate of improvement in middle-aged mortality disease-specific mortality rates for the most was not anticipated, with serious consequences for prevalent causes of death, which together account the financing of pension arrangements. In the UK, for some two-thirds of all deaths. Between 1971 and the market for annuities collapsed, employers closed 2001 their mortality rates tended to fall by defined benefit pension schemes for much of the proportionately similar amounts. The tendency was private sector and the age at which the state pension stronger for men than for women. Male mortality is paid was increased first by five years for women, to rates for all of these prevalent causes of death be followed by a further increase of three years for reduced by around 50-75%. For women, in contrast, all citizens. the fall in mortality rates was in the 50-75% range for Most of the actuarial and epidemiological research only five of the prevalent causes of death, while into the causes of this increasing longevity is disease- another reduced by around one-third and two specific, usually coronary heart disease (Unal, increased (it is tempting to attribute the rise in Critchley & Capewell, 2005; Bajekal et al., 2013) or women’s COPD and lung cancer mortality to their cerebrovascular disease (Raine et al., 2009); and tobacco smoking, although the fall in their CHD assumed that the main determinants of the falling mortality, which is also tobacco-related, gives pause mortality rates would be either change in risk factors for thought). Table 1. Age-standardised mortality rates per million population; England & Wales, 1971-2010; selected causes, by prevalence. Males Females 1971 2010 Percentage 1971 2010 Percentage change change 1971-2010 1971-2010 Ischaemic heart disease 3,801 1,083 -71.5 1,668 478 -71.3 Cerebrovascular disease 1,541 422 -72.6 1,352 396 -70.7 Chronic obstructive 944 314 -66.7 193 212 +9.8 pulmonary disease Pneumonia 920 260 -71.7 624 212 -66.0 Carcinoma of lung 1,066 465 -56.4 183 299 +63.4 Accidents and violence 333 170 -48.9 166 53 -68.1 Carcinoma of breast 4 2 -50.0 379 245 -35.4 Carcinoma of stomach 317 72 -77.3 149 33 -77.9 Source: Office for National Statistics (2011). 166 COMMENT AND DEBATE Something similar happened in the late Figure 1 shows the change in the death rates of nineteenth and early twentieth centuries when people aged 65 years in England and Wales from mortality due to most of the prevalent infectious the 1840s to 2010. The mortality rates of women diseases of childhood fell at around the same time and men changed little (if anything they increased) (McKeown & Lowe, 1966; McKeown, 1979), during the nineteenth century from 1841-1849 to although in that instance the situation was simpler 1891-1899. After 1900 women’s mortality rate at conceptually because infectious disease age 65 years fell in an approximately linear fashion epidemiology already contained the idea of host across the whole of the twentieth century. Men’s resistance. Positive health can be seen as the mortality in contrast, after an initial fall, plateaued chronic degenerative disease equivalent of host from 1931-39 to 1961-69 followed by a precipitous resistance. fall to near-equality with the women’s rate by 2011. Figure (. Death rates of men and women at age1 years per 455 live population of that age in England and Wales to Source: Human Mortality Database. The century-long, near linear fall in women’s requires explanation (the plateau coincided with mortality may have received too little attention. It the introduction of antibiotics and widened access suggests a cumulative process driven by rising living to, and sizeable investment in, medical care), with standards, with little sign of any large impact from the latter a process of catching up for lost time. the introduction of antibiotics, the start of the If it is accepted that risk factor change and National Health Service or the changing prevalence medical innovation have difficulty, on their own, in of tobacco smoking; also, it is easier to predict accounting for the twentieth century’s near linear future developments, including the necessary level fall in the mortality of 65 year old women and the of pension contributions, from linear change. proportionately similar fall in mortality from most Prediction would have been more difficult from the of the prevalent causes of death, then it may be change in men’s mortality at age 65 years: first the worth considering positive health as a third factor in three-decade plateau from the 1930s to the 1960s the explanation of longevity risk. The next section of suggested stability rather than change, then the the present paper attempts to do so by examining subsequent precipitous decline suggested rapid the biological and social life course of the golden change. On one reading, it is the former that cohort, operationalised as those currently aged 85 167 COMMENT AND DEBATE years. In both cases, biological and social, the Biological Life Course various life course stages of these 85-year-old Figure 2 presents a model of the biological life persons will be located within their UK social history course. The model was described by Strachan and context. Sheikh (2004) in relation to lung function; here it is generalised to positive health. Figure 2. Source: based on Strachan and Sheikh (2004), own visualization. First there is a phase of growth and and residential crowding, is associated with late development, during which an initial fertilised cell adolescent sub-clinical abnormality (erythrocyte divides and replicates to an estimated 37.2 trillion sedimentation rate, proteinuria, blood pressure, cells by around age 20 years (Bianconi et al., 2003). body mass index, vulnerability to stress) which, in Functional capacity subsequently is lost: slowly at turn, is associated with premature death in middle first, when the loss is more than compensated by age (Sundin, Udumyan, Sjostrom & Montgomery, experience, anticipation and conditioned reflexes, 2014; Bergh et al., 2014). Early death also results as shown by professional footballers and Olympic when normal growth and development is followed track and field champions; then a more rapid by adverse circumstances which accelerate the rate decline to morbidity and eventual death (full line in of functional loss (dashed line in figure 2); for figure 2).